Trait association studies of grain yield with its related attributes in rice (Oryza sativa L.)

Introduction

Rice is one of the most important cereal crops in the world, serving as the staple food for more than half of the global population and playing a vital role in food and nutritional security, particularly in Asia (FAO, 2025) [6] . According to the FAO (2025) [6] , rice is cultivated globally on approximately 165 million hectares, producing about 543 million tonnes (milled basis) with an average productivity of approximately 4.7 t ha⁻¹. Since grain yield is a complex quantitative trait controlled by the interaction of several component characters, an understanding of the relationships among these traits is essential for developing efficient selection strategies (Falconer and Mackay, 1996; Singh and Chaudhary, 1985) [5,15] . Moreover, these yield-contributing traits are governed by multiple genes and their interactions with the environment, making the study of their association with grain yield indispensable for improving selection efficiency in rice breeding programmes (Wright, 1921; Barde et al., 2024) [18,1] . Character association analysis helps identify the key traits that contribute to yield and provides a scientific basis for selecting desirable genotypes. It provides information on the degree of association between traits however it does not distinguish whether these relationships are due to direct effects or are mediated through other traits. Path coefficient analysis, originally proposed by Wright (1921) [18] and later applied to plant breeding by Dewey and Lu (1959) [2] , partitions the correlation coefficients into direct and indirect effects, thereby providing a clearer understanding of the contribution of individual traits to grain yield. The combined use of correlation and path coefficient analyses enables breeders to identify the most influential yield-contributing traits and develop efficient selection criteria for crop improvement. Therefore, greater emphasis can be placed on those traits that show a strong association with grain yield, thereby improving the efficiency of selection in rice breeding programmes. With this background, the present study was undertaken to investigate the character associations and path coefficients among grain yield and its component traits in rice genotypes evaluated across multiple environments.

The findings are expected to provide valuable information for identifying key traits that can be effectively utilized in breeding programmes aimed at improving grain yield.

Materials and Methods

To examine the correlation and path coefficient relationships among yield and its related traits, a total of 28 rice genotypes comprising of 25 lines and 3 checks were evaluated (Table -1). The experiment was carried out during the kharif, 2025 at three locations: Agricultural Research Station, Kampasagar (STZ), Regional Agricultural Research Station, Warangal (CTZ) and Agricultural Research Station, Kunaram (NTZ). These three experimental locations represent the major agro-climatic zones of Telangana, namely the Northern Telangana Zone (NTZ), Central Telangana Zone (CTZ) and Southern Telangana Zone (STZ). The experiment was conducted using a Randomized Block Design (RBD) with three replications at each location. Each experimental plot measured 4.5 m², and the 28 rice genotypes were transplanted by maintaining a spacing of 20 cm between rows and 15 cm between plants. Normal agronomical practices and plant protection measures with external inputs such as supplementary irrigation and fertilizers were given at appropriate time to maintain a consistent and healthy crop stand. Based on the data recorded from all 28 genotypes, correlation and path coefficient analyses were performed to assess the relationships among grain yield and its associated traits. Correlation coefficients among the studied traits were estimated following the procedure described by Singh and Chaudhary (1994). The significance of the correlation coefficients was tested by comparing the calculated values with the corresponding tabulated values at (n-1) degrees of freedom at the 5% and 1% probability levels. Path coefficient analysis was performed according to the method originally proposed by Wright (1921) [18] and subsequently elaborated by Dewey and Lu (1959) [2] . This analysis was used to partition the correlation coefficients into direct and indirect effects, thereby identifying the traits that contributed most effectively to grain yield.

Results and Discussion

A clear understanding of the relationships among yield and its component traits is essential for developing effective selection criteria in crop improvement programmes.

Although correlation analysis provides information on the association between traits, it does not clearly distinguish whether a trait influences grain yield directly or through its association with other traits. Therefore, relying solely on correlation coefficients may sometimes lead to misleading conclusions and reduce the efficiency of indirect selection. To overcome this limitation, path coefficient analysis, as proposed by Wright (1921) [18] partitions the total correlation into direct and indirect effects. This approach provides a better understanding of the contribution of individual traits to grain yield, especially when several traits simultaneously influence yield. Such information is valuable for identifying the most important traits to be considered in selection programmes aimed at improving grain yield.

Correlation Studies

Based on the observations recorded from 28 rice genotypes, character association analysis was carried out to assess the relationships between grain yield and its component traits. The analysis was performed to identify the traits associated with higher grain yield and to understand the factors contributing to improved yield performance in the hybrids. The results of the correlation analysis are presented in Table 2. The correlation analysis over three locations revealed distinct relationships among grain yield and its component traits (Table X). Among the traits evaluated, the number of filled grains per panicle recorded the highest positive correlation with grain yield (r = 0.8431**) followed by the number of productive tillers per plant (r = 0.8348**). This suggests that both these traits play a major role in determining grain yield. Panicle length also showed a strong positive association with grain yield (r = 0.6340**) indicating that genotypes with longer panicles generally possess greater yield potential. Similarly, test weight exhibited a significant positive correlation (r = 0.4860**) highlighting the contribution of grain weight to overall productivity. Similar positive associations between these yield-contributing traits and grain yield have been reported by Kanwar and Nag (2019), Saleh et al. (2020) [13] , Naik et al. (2021) [11] , Barde et al. (2024) [1] , and Kumar et al. (2025). These positive associations indicate that improvements in these traits are likely to result in higher grain yield and, therefore, they may serve as effective selection criteria in rice breeding programmes. Plant height (r = 0.1156), kernel length (r = 0.2455) and kernel breadth (r = 0.2576) exhibited weak positive but nonsignificant correlations with grain yield, indicating that these traits had limited direct influence on yield improvement. Similar positive associations between plant height and grain yield were reported by Kafi et al. (2021) [8] , Vennela et al. (2021) [17] and Shanmugam et al. (2024) [14] , whereas Manivelan et al. (2022) [10] observed positive associations of kernel length with grain yield, and Vennela et al. (2021) [17] also reported a positive association of kernel breadth with grain yield. Similarly, the length-to-breadth ratio (L/B) was negatively associated with grain yield (r = -0.1988), suggesting that grain shape had little contribution to yield performance under the present study. A contrasting positive association of kernel L/B ratio with grain yield was reported by Manivelan et al. (2022) [10] , indicating that the contribution of grain shape to yield may vary with genetic background and environmental conditions.

In contrast, days to 50% flowering showed a negative nonsignificant association with grain yield (r = -0.4481) indicating that delayed flowering tended to reduce grain yield under the pooled environmental conditions. This negative association suggests that relatively early flowering genotypes may have an advantage in achieving higher productivity. while significant negative correlations were observed with number of productive tillers (r = -0.5753**) and number of filled grains per panicle (r = -0.4001*) suggesting that late-flowering genotypes generally produced fewer productive tillers and filled grains, ultimately leading to reduced grain yield. These findings are in agreement with the reports of Saleh et al. (2020) [13] , Sharma et al. (2024) [16] , and Prasad et al., who also observed negative associations between days to 50% flowering and major yield-contributing traits. The interrelationships among yield-contributing traits further revealed that panicle length was positively and significantly associated with number of productive tillers (r = 0.6356**) number of filled grains per panicle (r = 0.5866**) test weight (r = 0.5476**) kernel length (r = 0.4606*) and kernel breadth (r = 0.3878*) indicating that selection for increased panicle length may simultaneously improve other important yield-contributing traits. A very strong positive association was observed between number of productive tillers and number of filled grains per panicle (r = 0.8244**) suggesting that genotypes producing more productive tillers also tended to produce more filled grains, thereby contributing substantially to grain yield. Test weight also showed strong positive correlations with kernel length (r = 0.8996**) and kernel breadth (r = 0.8294**) reflecting the close relationship between grain size attributes and grain weight. Furthermore, kernel length and kernel breadth were positively associated (r=0.8105**) whereas the length-to-breadth ratio exhibited a strong negative correlation with kernel breadth (r = -0.7663**).

Path Coefficient Analysis

Path coefficient analysis was performed using pooled data from the three test locations to partition the correlation coefficients into direct and indirect effects of yieldattributing traits on grain yield (Table 3). Among the traits studied, number of productive tillers per plant exerted the highest positive direct effect on grain yield (0.4706) followed by number of filled grains per panicle (0.3205) kernel breadth (0.2694) length-to-breadth ratio (0.2092) and test weight (TW) (0.1699). The strong positive direct effects of number of productive tillers per plant and number of filled grains per panicle together with their significant positive correlations with grain yield, indicate that these traits are the major contributors to yield improvement. Therefore, direct selection for increased productive tillers and a higher number of filled grains per panicle would be effective for enhancing grain yield across different environments. Similar findings have been reported by Naik et al. (2021) [11] , Jangala et al. (2022) [7] , Barde et al. (2024) [1] , Dhruw et al. (2025) [3] , and Tiwari et al. (2025) [16] , who also identified these traits as major contributors to grain yield through their positive direct effects. Unlike the traits with positive direct effects, days to 50% flowering (-0.1034), plant height (-0.0215), panicle length (-0.0138) and kernel length (-0.1621) exhibited negative direct effects on grain yield. The negative direct effect of days to 50% flowering is partially consistent with the negative association between flowering time and grain yield reported by Sharma et al. (2024) [16] . However, the negative direct effects of plant height, panicle length and kernel length observed in the present study contrast with the findings of Vennela et al. (2021) [17] , Manivelan et al. (2022) [10] , Sadhana et al. (2022) [12] and Dora et al. (2025) [4] , who reported positive direct effects of these traits on grain yield. Although panicle length showed a negative direct effect (-0.0138) it maintained a strong positive correlation with grain yield (0.5792) largely because of the favourable indirect effect through number of productive tillers (0.2729) number of filled grains per panicle (0.2063) and test weight (0.0844). This indicates that the contribution of panicle length to grain yield is mainly mediated through its favourable association with other important yield components rather than through its own direct influence. Therefore, panicle length may be considered an indirect selection criterion for improving grain yield. The number of productive tillers per plant also contributed to grain yield indirectly by increasing the number of filled grains per panicle (0.2547) and test weight (0.0864) highlighting its important role in improving yield. Similarly, the number of filled grains per panicle had positive indirect effects through the number of productive tillers (0.3740) and test weight (0.0714) indicating that these traits work together to enhance grain yield. Test weight had a positive direct effect (0.1699) and also contributed indirectly through number of productive tillers (0.2394) and number of filled grains per panicle (0.1347). These findings indicate that heavier grains, together with improved tillering ability and grain filling, contribute substantially to enhanced grain yield. These results are consistent with the findings of Manivelan et al. (2022) [10] and Dora et al. (2025) [4] . Kernel breadth exhibited a relatively high positive direct effect (0.2694) but its overall association with grain yield remained comparatively low (0.1320) suggesting that several unfavourable indirect effects, particularly through kernel length (-0.1231) and length-to-breadth ratio (-0.1520) reduced its overall contribution. Similarly, the length-tobreadth ratio recorded a positive direct effect (0.2092) but its overall correlation with grain yield was negligible (0.0021) indicating that its positive direct effect was reduced by negative indirect effects through kernel breadth (-0.1958) and test weight (-0.0436).

Conclusion

Overall, the pooled correlation and path coefficient analyses consistently identified the number of productive tillers per plant, number of filled grains per panicle and test weight as the major contributors to grain yield across the three environments. These traits not only exhibited strong positive associations with grain yield but also exerted favourable direct and indirect effects, highlighting their importance in determining yield performance. Although panicle length showed a positive association with grain yield, its contribution was mainly through its indirect influence on productive tillers and filled grains per panicle. Therefore, simultaneous selection for productive tillers, filled grains per panicle, test weight and panicle length is likely to be the most effective approach for improving grain yield and developing high-yielding rice genotypes with broad adaptation across diverse environments.

Acknowledgement

AcknowledgementI gratefully acknowledge Professor Jayashankar Telangana Agricultural University for the financial support that enabled this research.

References

1. Barde YD, Wandhare MR, Madke VS, Manapure P, Bagade VH, Kunte SG, Pardhi KC. Correlation and path analysis studies in rice (Oryza sativa L.). International Journal of Advanced Biochemistry Research. 2024;8:973-979.
2. Dewey DR, Lu KH. A correlation and path-coefficient analysis of components of crested wheatgrass seed production. Agronomy Journal. 1959;51:515-518.
3. Dhruw K, Gauraha D, Sao A. Assessment of genetic divergence and character association studies on germplasm accessions of rice (Oryza sativa L.). Journal of Advances in Biology and Biotechnology. 2025;28(1):719-727.
4. Dora SJ, Harini AS, Senguttuvel P, Sanjeeva Rao D. Path analysis and correlation studies for identifying superior restorer lines in hybrid rice breeding. International Journal of Advanced Biochemistry Research. 2025;9(12):639-645.
5. Falconer DS, Mackay TFC. Introduction to Quantitative Genetics. 4th ed. Harlow: Longman; 1996.
6. FAO. World Food and Agriculture: Statistical Yearbook 2025. Rome: Food and Agriculture Organization of the United Nations; 2025.
7. Jangala DJ, Amudha K, Geetha S, Uma D. Studies on genetic diversity, correlation and path analysis in rice germplasm. Electronic Journal of Plant Breeding. 2022;13(2):655-662.
8. Kafi SH, Abiodun EA, Bunmi O, Kyung-Ho K. Correlation coefficient and path analyses of yield and yield related traits of Korean double haploid rice for germplasm improvement in Nigeria. American Journal of Agriculture and Forestry. 2021;9(3):114-121.
9. Kumar D, Vimal SC, Bisen P, Kasaudhan SK, Kushwaha S, Prajapati DK. Genetic analysis of grain yield and its associated traits in rice (Oryza sativa L.). Journal of Advances in Biology and Biotechnology. 2025;28(4):840-846.
10. Manivelan K, Hepziba JS, Suresh R, Theradimani M, Renuka R, Gnanamalar RP. Inherent variability, correlation and path analysis in lowland rice (Oryza sativa L.). Biological Forum -An International Journal. 2022;14(2):771-778.
11. Naik MVK, Pillai MA, Saravanan S. Assessment of correlation and path coefficient analysis for yield attributing and quality traits in promising rice varieties cultivated in Tamil Nadu. The Journal of Phytopharmacology. 2021;10(2):139-143.
12. Sadhana P, Raju CD, Subba Rao LV, Kuna A. Studies on variability, correlation and path coefficient analysis for yield and quality traits in rice (Oryza sativa L.) genotypes. Electronic Journal of Plant Breeding. 2022;13(2):670-678.
13. Saleh MM, Salem KFM, Elabd AB. Definition of selection criterion using correlation and path coefficient analysis in rice (Oryza sativa L.) genotypes. Bulletin of the National Research Centre. 2020;44:143-148.
14. Shanmugam A, Suresh R, Ramanathan A, Anandhi P, Pushpa R, Sassikumar D. Correlation and path coefficient analysis among yield and yield attributing traits of rice landraces. Current Innovation in Agriculture Science. 2024;1(1):35-42.
15. Singh RK, Chaudhary BD. Biometrical Methods in Quantitative Genetic Analysis. New Delhi: Kalyani Publishers; 1985.
16. Tiwari R, Sharma B, Parikh M, Gauraha D. Characters association and selection criteria for irrigated ecosystem in rice (Oryza sativa L.) genotypes. Journal of Scientific Research and Reports. 2025;31(1):723-731.
17. Vennela M, Srinivas B, Ram Reddy V, Balram N. Studies on correlation and path coefficient analysis in hybrid rice (Oryza sativa L.) for yield and quality traits. International Journal of Bio-resource and Stress Management. 2021;12(5):496-505.
18. Wright S. Correlation and causation. Journal of Agricultural Research. 1921;20:557-585.